Ureader.com  
Microsoft software help and Community
   home   |   control panel login   |   archive   |  
 
Exchange
2000.active.directory
2000.admin
2000.announcements
2000.app.conversion
2000.applications
2000.clients
2000.clustering
2000.connectivity
2000.development
2000.documentation
2000.general
2000.information.store
2000.interop
2000.kms
2000.misc
2000.protocols
2000.realtime.collabo.
2000.setup
2000.transport
2000.win2000
admin
application.conversion
applications
clients
clustering
connectivity
design
development
misc
mobility
setup
tools
  
 
date: Wed, 5 Mar 2008 15:58:01 -0800,    group: microsoft.public.exchange.clustering        back       


stop/start CCR CMS   
I have a 2007 sp1 CCR cluster.  If I stop the active node by using either 
cmdlet or EMC 

Q1: will it automatically failover to the passive node?  I think it will.

Q2: after maintenance and starting original active node (was stopped early) 
will the cluster failback?  Or can I started it since there will be two 
active nodes.

Thanks.
date: Wed, 5 Mar 2008 15:58:01 -0800   author:   Chris

Re: stop/start CCR CMS   
Hi Chris:

A1: No, it will not.  Failover implies an unscheduled outage.  A move, or 
what we also call a handoff, is a scheduled outage.  So if you need to 
perform maintenance, instead of stopping the clustered mailbox server, you 
move the clustered mailbox server to the passive node, thereby making it the 
new active node.

A2: When you're done with the maintenance, you move the clustered mailbox 
server back to the node, and then if necessary perform the same maintenance 
on the other node.

Hope this helps.
-- 
Regards,

Scott Schnoll
Microsoft Corporation
This posting is provided "AS IS" with no warranties, and confers no
rights. Please do not send email directly to this alias. This alias is for
newsgroup purposes only.


"Chris"  wrote in message 
news:1B086D7F-49ED-492B-B0D7-2DB7AB27D920@microsoft.com...
>I have a 2007 sp1 CCR cluster.  If I stop the active node by using either
> cmdlet or EMC
>
> Q1: will it automatically failover to the passive node?  I think it will.
>
> Q2: after maintenance and starting original active node (was stopped 
> early)
> will the cluster failback?  Or can I started it since there will be two
> active nodes.
>
> Thanks.
date: Wed, 5 Mar 2008 16:42:14 -0800   author:   Scott Schnoll [MSFT]

Re: stop/start CCR CMS   
Scott,
  Thanks for your reply.  I have a few more questions.

1. When or why should I stop cluster node?  I'm not quite understand its 
purpose.  I think it's new to CCR.

2. If I moved an active node to passive node, then stopped cluster on active 
node. I guess this will make the node "offline" or not available for the 
cluster?  If I start the node again does it resume its previous state, i.e. 
passive node?

3. I think CCR is using the majority node concept.  With just two nodes if 
one is stopped is 50% a majority?  Is witness also considered as a "node"?

thanks.

"Scott Schnoll [MSFT]" wrote:

> Hi Chris:
> 
> A1: No, it will not.  Failover implies an unscheduled outage.  A move, or 
> what we also call a handoff, is a scheduled outage.  So if you need to 
> perform maintenance, instead of stopping the clustered mailbox server, you 
> move the clustered mailbox server to the passive node, thereby making it the 
> new active node.
> 
> A2: When you're done with the maintenance, you move the clustered mailbox 
> server back to the node, and then if necessary perform the same maintenance 
> on the other node.
> 
> Hope this helps.
> -- 
> Regards,
> 
> Scott Schnoll
> Microsoft Corporation
> This posting is provided "AS IS" with no warranties, and confers no
> rights. Please do not send email directly to this alias. This alias is for
> newsgroup purposes only.
> 
> 
> "Chris"  wrote in message 
> news:1B086D7F-49ED-492B-B0D7-2DB7AB27D920@microsoft.com...
> >I have a 2007 sp1 CCR cluster.  If I stop the active node by using either
> > cmdlet or EMC
> >
> > Q1: will it automatically failover to the passive node?  I think it will.
> >
> > Q2: after maintenance and starting original active node (was stopped 
> > early)
> > will the cluster failback?  Or can I started it since there will be two
> > active nodes.
> >
> > Thanks. 
> 
>
date: Thu, 6 Mar 2008 07:50:03 -0800   author:   Chris

Re: stop/start CCR CMS   
Stopping and starting a clustered mailbox server (or shutting down or 
rebooting a node hosting a clustered mailbox server) is not new to CCR, or 
unique to Exchange 2007.  It's just part of possible maintenance you might 
need to perform.  See 
http://technet.microsoft.com/en-us/library/aa997676(EXCHG.80).aspx.

Say you have two nodes: NODEA and NODEB.  NODEA is currently hosting the 
clustered mailbox server (CMS).  If you need to perform maintenance on 
NODEA, you (1) move the CMS to NODEB, and if necessary, also move the 
Cluster Group from NODEA to NODEB.  At this point, NODEA becomes your 
passive, and NODEB becomes your active.  When you restart NODEA, it will 
remain the passive node (provided that you have not modified the default 
cluster policy settings, such as preferred owner or failback settings).

CCR uses a variant of an MNS quorum called a Majority Node Set with File 
Share Witness.  In this model, you have two nodes, plus a third quorum voter 
which is implemented as a file share on a separate system.  The witness is 
not considered a node because the term node refers to a member of the 
cluster.  The server hosting the file share for the file share witness is 
not part of the cluster, but rather, a separate server entirely.  In fact, 
this was one of reasons we came up with the file share witness; we wanted 
CCR to be two cluster nodes only but still have a way to maintain quorum 
without using a shared disk.  The file share witness feature was created so 
that you don't need a third node in your cluster. See 
http://technet.microsoft.com/en-us/library/bb124521(EXCHG.80).aspx for more 
information.

-- 
Regards,

Scott Schnoll
Microsoft Corporation
This posting is provided "AS IS" with no warranties, and confers no
rights. Please do not send email directly to this alias. This alias is for
newsgroup purposes only.


"Chris"  wrote in message 
news:793DA15B-883F-4092-A5A7-B7F3FD3BF5F1@microsoft.com...
> Scott,
>  Thanks for your reply.  I have a few more questions.
>
> 1. When or why should I stop cluster node?  I'm not quite understand its
> purpose.  I think it's new to CCR.
>
> 2. If I moved an active node to passive node, then stopped cluster on 
> active
> node. I guess this will make the node "offline" or not available for the
> cluster?  If I start the node again does it resume its previous state, 
> i.e.
> passive node?
>
> 3. I think CCR is using the majority node concept.  With just two nodes if
> one is stopped is 50% a majority?  Is witness also considered as a "node"?
>
> thanks.
>
> "Scott Schnoll [MSFT]" wrote:
>
>> Hi Chris:
>>
>> A1: No, it will not.  Failover implies an unscheduled outage.  A move, or
>> what we also call a handoff, is a scheduled outage.  So if you need to
>> perform maintenance, instead of stopping the clustered mailbox server, 
>> you
>> move the clustered mailbox server to the passive node, thereby making it 
>> the
>> new active node.
>>
>> A2: When you're done with the maintenance, you move the clustered mailbox
>> server back to the node, and then if necessary perform the same 
>> maintenance
>> on the other node.
>>
>> Hope this helps.
>> -- 
>> Regards,
>>
>> Scott Schnoll
>> Microsoft Corporation
>> This posting is provided "AS IS" with no warranties, and confers no
>> rights. Please do not send email directly to this alias. This alias is 
>> for
>> newsgroup purposes only.
>>
>>
>> "Chris"  wrote in message
>> news:1B086D7F-49ED-492B-B0D7-2DB7AB27D920@microsoft.com...
>> >I have a 2007 sp1 CCR cluster.  If I stop the active node by using 
>> >either
>> > cmdlet or EMC
>> >
>> > Q1: will it automatically failover to the passive node?  I think it 
>> > will.
>> >
>> > Q2: after maintenance and starting original active node (was stopped
>> > early)
>> > will the cluster failback?  Or can I started it since there will be two
>> > active nodes.
>> >
>> > Thanks.
>>
>>
date: Thu, 6 Mar 2008 08:50:14 -0800   author:   Scott Schnoll [MSFT]

Google
 
Web ureader.com


    COPYRIGHT 2007, YARDI TECHNOLOGY LIMITED, ALL RIGHT RESERVE  |   contact us